| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 13, | |
| "global_step": 49, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02040816326530612, | |
| "grad_norm": 0.7881951332092285, | |
| "learning_rate": 2e-05, | |
| "loss": 2.7509, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02040816326530612, | |
| "eval_loss": 2.6902382373809814, | |
| "eval_runtime": 269.5606, | |
| "eval_samples_per_second": 6.288, | |
| "eval_steps_per_second": 3.146, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04081632653061224, | |
| "grad_norm": 0.789082407951355, | |
| "learning_rate": 4e-05, | |
| "loss": 2.7449, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.061224489795918366, | |
| "grad_norm": 0.7354114055633545, | |
| "learning_rate": 6e-05, | |
| "loss": 2.7164, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.08163265306122448, | |
| "grad_norm": 0.7292255759239197, | |
| "learning_rate": 8e-05, | |
| "loss": 2.7174, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 0.6898028254508972, | |
| "learning_rate": 0.0001, | |
| "loss": 2.6891, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.12244897959183673, | |
| "grad_norm": 0.6861400604248047, | |
| "learning_rate": 0.00012, | |
| "loss": 2.6545, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 0.7510350346565247, | |
| "learning_rate": 0.00014, | |
| "loss": 2.5656, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 0.8011165261268616, | |
| "learning_rate": 0.00016, | |
| "loss": 2.4519, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1836734693877551, | |
| "grad_norm": 0.8624005317687988, | |
| "learning_rate": 0.00018, | |
| "loss": 2.3178, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 0.8004987835884094, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1783, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22448979591836735, | |
| "grad_norm": 0.6362400054931641, | |
| "learning_rate": 0.000199985736255971, | |
| "loss": 2.0252, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.24489795918367346, | |
| "grad_norm": 0.7930936217308044, | |
| "learning_rate": 0.0001999429490929718, | |
| "loss": 1.8839, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2653061224489796, | |
| "grad_norm": 0.5149843096733093, | |
| "learning_rate": 0.00019987165071710527, | |
| "loss": 1.8064, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2653061224489796, | |
| "eval_loss": 1.6734941005706787, | |
| "eval_runtime": 271.2615, | |
| "eval_samples_per_second": 6.249, | |
| "eval_steps_per_second": 3.126, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.42121434211730957, | |
| "learning_rate": 0.00019977186146800707, | |
| "loss": 1.7922, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 0.3523242771625519, | |
| "learning_rate": 0.0001996436098130433, | |
| "loss": 1.7711, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 0.3384595215320587, | |
| "learning_rate": 0.00019948693233918952, | |
| "loss": 1.7152, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3469387755102041, | |
| "grad_norm": 0.34942421317100525, | |
| "learning_rate": 0.00019930187374259337, | |
| "loss": 1.7112, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3673469387755102, | |
| "grad_norm": 0.31712639331817627, | |
| "learning_rate": 0.00019908848681582391, | |
| "loss": 1.7059, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3877551020408163, | |
| "grad_norm": 0.2875436842441559, | |
| "learning_rate": 0.00019884683243281116, | |
| "loss": 1.6468, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 0.24433130025863647, | |
| "learning_rate": 0.00019857697953148037, | |
| "loss": 1.6408, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.21414674818515778, | |
| "learning_rate": 0.00019827900509408581, | |
| "loss": 1.616, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.4489795918367347, | |
| "grad_norm": 0.21537622809410095, | |
| "learning_rate": 0.00019795299412524945, | |
| "loss": 1.609, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.46938775510204084, | |
| "grad_norm": 0.2432074397802353, | |
| "learning_rate": 0.00019759903962771156, | |
| "loss": 1.6066, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4897959183673469, | |
| "grad_norm": 0.2359839379787445, | |
| "learning_rate": 0.00019721724257579907, | |
| "loss": 1.5851, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 0.22065888345241547, | |
| "learning_rate": 0.00019680771188662044, | |
| "loss": 1.5739, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5306122448979592, | |
| "grad_norm": 0.20339132845401764, | |
| "learning_rate": 0.0001963705643889941, | |
| "loss": 1.5513, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5306122448979592, | |
| "eval_loss": 1.4832030534744263, | |
| "eval_runtime": 271.2449, | |
| "eval_samples_per_second": 6.249, | |
| "eval_steps_per_second": 3.126, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5510204081632653, | |
| "grad_norm": 0.18875224888324738, | |
| "learning_rate": 0.00019590592479012023, | |
| "loss": 1.5378, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.18564417958259583, | |
| "learning_rate": 0.00019541392564000488, | |
| "loss": 1.5212, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5918367346938775, | |
| "grad_norm": 0.16226942837238312, | |
| "learning_rate": 0.00019489470729364692, | |
| "loss": 1.5391, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 0.15650039911270142, | |
| "learning_rate": 0.00019434841787099803, | |
| "loss": 1.511, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6326530612244898, | |
| "grad_norm": 0.15976540744304657, | |
| "learning_rate": 0.00019377521321470805, | |
| "loss": 1.5119, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6530612244897959, | |
| "grad_norm": 0.16409288346767426, | |
| "learning_rate": 0.00019317525684566685, | |
| "loss": 1.4909, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.673469387755102, | |
| "grad_norm": 0.15468019247055054, | |
| "learning_rate": 0.00019254871991635598, | |
| "loss": 1.4951, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6938775510204082, | |
| "grad_norm": 0.1462036371231079, | |
| "learning_rate": 0.00019189578116202307, | |
| "loss": 1.4643, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.1541963368654251, | |
| "learning_rate": 0.00019121662684969335, | |
| "loss": 1.5159, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7346938775510204, | |
| "grad_norm": 0.14798064529895782, | |
| "learning_rate": 0.00019051145072503215, | |
| "loss": 1.4741, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7551020408163265, | |
| "grad_norm": 0.13914817571640015, | |
| "learning_rate": 0.00018978045395707418, | |
| "loss": 1.4788, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.7755102040816326, | |
| "grad_norm": 0.15608824789524078, | |
| "learning_rate": 0.00018902384508083517, | |
| "loss": 1.4687, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7959183673469388, | |
| "grad_norm": 0.14460116624832153, | |
| "learning_rate": 0.00018824183993782192, | |
| "loss": 1.482, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7959183673469388, | |
| "eval_loss": 1.411073088645935, | |
| "eval_runtime": 271.292, | |
| "eval_samples_per_second": 6.248, | |
| "eval_steps_per_second": 3.126, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 0.15740551054477692, | |
| "learning_rate": 0.00018743466161445823, | |
| "loss": 1.4486, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8367346938775511, | |
| "grad_norm": 0.14149661362171173, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 1.4353, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.14034292101860046, | |
| "learning_rate": 0.0001857457136130651, | |
| "loss": 1.4523, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.8775510204081632, | |
| "grad_norm": 0.1487722396850586, | |
| "learning_rate": 0.00018486442574947511, | |
| "loss": 1.4095, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.8979591836734694, | |
| "grad_norm": 0.17400234937667847, | |
| "learning_rate": 0.00018395892819696389, | |
| "loss": 1.4414, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 0.1741325408220291, | |
| "learning_rate": 0.00018302947927123766, | |
| "loss": 1.4379, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9387755102040817, | |
| "grad_norm": 0.15319454669952393, | |
| "learning_rate": 0.00018207634412072764, | |
| "loss": 1.405, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.9591836734693877, | |
| "grad_norm": 0.15876264870166779, | |
| "learning_rate": 0.00018109979465095013, | |
| "loss": 1.4122, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.9795918367346939, | |
| "grad_norm": 0.17120805382728577, | |
| "learning_rate": 0.00018010010944693848, | |
| "loss": 1.4132, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.1436116099357605, | |
| "learning_rate": 0.00017907757369376985, | |
| "loss": 1.416, | |
| "step": 49 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 196, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 49, | |
| "total_flos": 1.0209350802997248e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |